P-Amyloid Precursor Epitopes in Muscle Fibers of Inclusion Body Myositis

Abstract

Sporadic inclusion body myositis (IBM) and hereditary inclusion body myopathy (hIBM) are severe and progressive muscle diseases, characterized pathologically by vacuolated muscle fibers that contain 15-to 21-nm cytoplasmic tubulofilaments (CTFs). Those vacuolated muscle fibers also contain abnormally accumulated ubiquitin and P-amyloid protein (AP), and they contain amyloid in 0-pleated sheets as indicated by Congo red and crystal violet positivity. Using several well-characterized antibodies, we have now demonstrated that, in addition to AP, two other epitopes, N-terminal and C-terminal, of the P-amyloid precursor protein (PPP) are abnormally accumulated in IBM vacuolated muscle fibers and similarly in hlBM. At the light microscopy level, immunoreactivities of N-and C-epitopes of PPP closely colocalized with AP and ubiquitin immunoreactivities. However, by immunogold electronmicroscopy, even though N-, C-, and AQ epitopes of PPP and ubiquitin colocalized at the amorphous and dense floccular structures, only AP was localized to the 6to 10-nm amyloid-like fibrils and only ubiquitin was localized to CTFs. PPP immunoreactive structures were often in proximity to CTFs, but CTFs themselves never contained PPP immunoreactivities. The fact that AP but not C-or N-terminal epitopes of PPP localized to the 6to 10-nm amyloid-like fibrils suggests that free AP might be generated during PPP processing and, after aggregation, may be responsible for the amyloid present within IBM muscle fibers. Our study demonstrates that three epitopes of PPP accumulate abnormally in diseased human muscle, and therefore this phenomenon is not unique to Alzheimer's disease, Down's syndrome brain, and Dutch-type cerebrovascular amyloidosis.

SECTION

of the much larger P-amyloid precursor protein (PPP) 124-261. Abnormal accumulation of AP in AD brain is thought to result from abnormal proteolytic cleavage of PPP, which produces the amyloidogenic fragment AP (reviewed in 127, 28)). In the dystrophic neurites of senile plaques of AD brain, in addition to AP there are abnormally accumulated C-terminal and Nterminal sequences of PPP 129 -38) . In respect to sporadic and hereditary IBM, we now have asked whether other epitopes of PPP, in addition to A@, are abnormally accumulated in IBM and, if so, how they relate ultrastructurally to AP and Ub immunoreacrivities. Using both light and EM immunocytochemistry, we localized three sequences of PPP, viz, the N-terminal 45-62, C-terminal 676-695, and AP sequence, and coloched them with each other and with Ub.

Material And Methods Patients

Immunolocalizations were performed on sections of diagnostic muscle biopsies (obtained with informal consent), from 33 patients, ages 5 to 76 years, with the following diagnoses: sporadic IBM, 14; autosomal-recessive hereditary JBM in an Iranian Jew, 1; polymyositis, 9; Duchenne muscular dystrophy, 1 ; amyotrophic lateral sclerosis, 4; normal muscle, 5. and CTFs by electronmicroscopy (Fig 1 ) . The majority of CTFs appeared as twisted helical filaments, having twist repeats of 41 to 5 5 nm. Some patients had clusters of densely packed small fibrils, 6 to 10 nm in diameter, which frequently were adjacent 10 CTFs (see Fig 1) . All patients with sporadic IBM, but not the one with the hereditary IBM, had crystal violet-positive (metachromatic red) and Congo red-positive amyloid inclusions in vacuolated muscle fibers (see Fig 1) .

Light Microscopic Immtllzocytocbemisty

Immunocytochemical stainings were performed on 10-ym transverse sections of fresh-frozen muscle biopsies, using peroxidase-antiperoxidase (PAP) and fluorescence methods, following the same general procedures as described E13, 14, 16) . Six well-characterized antibodies against PPP sequences were used (details in Table 1 ). These antibodies were shown to be very specific in our previous studies [15, 16, 431 . In addition, the following two commercially available antibodies against PPP sequences (Boehringer) were used: (1) rabbit polyclonal against sequence 1-40 of A@, diluted 1:20 to 1:40, and ( 2 ) mouse monoclonal, clone 22Cl1, that recognized sequence 60-100 of N-PPP diluted 1 : 20 to 1 : 40. Ub was localized with a monoclonal antibody, clone 042691GS (Chemicon, Temecula, CA) diluted 1:50 to 1:200 depending on the lot. In our previous study, this antibody proved to be specific and produced the same results as several other well-characterized monoclonal and polyclonal anti-Ub antibodies 114- 16, 43, 441. For PAP reactions, sections were preincubated for 60 min-utes in either normal goat or rabbit serum dilured 1 : 10, followed by 48 hours of incubation at 4°C in the appropriately diluted primary antibody, and then by 1 hour of incubation in a secondary antiserum. This was followed by a PAP complex, visualized by the DAB reaction. For fluorescence staining, the procedure was essentially the same as for PAP except that the secondary antibodies were labeled either with fluorescein isothiocyanate (FITC) or rhodamine. In some instances, the secondary antiserum was biotinylated and the reaction product was visualized by streptavidin-Texas red. Double immunolocalization of PPP epitopes with each other and with Ub was performed with double fluorescence labeling using FITC, Texas red, or rhodamine as described 116, 43, 443.

Specificity Of Ppp Immunoreactivities Was Determined By

( 1 ) omitting the primary antibody, (2) replacing the primary antibody with nonimmune serum, ( 3 ) absorbing the primary monoclonal antibody G-OP-1 with synthetic AP peptide sequence 8-17, and ( 4 ) absorbing primary polyclonal antiserum C8 with synthetic peptide 676-695.

Gold Immuno-Electronmicroscopy

PPP sequences and U b were localized on 10-pm unfixed frozen sections adhered to the bottom of 35-mm Petri dishes I1 3, 161. For single immunolocalization, sections were incubated in a primary antibody for 48 hours at 4"C, and then incubated in diluted secondary antibody conjugated to 10-nm gold particles (Amersham-Janssen Auro Probe EM) for 69 to 72 hours at 4°C. The sections were then fixed in a 2% paraformaldehyde 1.2% glutaraldehyde mixture, postfixed in osmium, and embedded in situ in the Petri dish, according to our method for cultured muscle 1451. The embedded section in the dish was viewed under phase-contrast microscopy and compared with an adjacent cross section that had been incubated with the same antibody as for EM but stained with the PAP reaction. The exact same muscle fibers that contained PAP-positive inclusions were identified in the adjacent gold-labeled Epon-embedded section, marked, and l-mmdiameter cores drilled out E45, 461. From the cores, each containing at least one vacuolated muscle fiber, thin sections were cut, counterstained with uranyl acetate and lead citrate, and examined by EM. This method has proved reliable for ultrastructural identification of CTFs in IBM vacuolated muscle fibers 147) and for ultrastructural immunogold labeling 113, 16, 48) .

For ultrastructural double immunolocalization, the method was essentially the same, except that sections were incubated simultaneously in two antibodies directed against different antigens and raised in different species. This was followed by incubation in two correspondingly appropriate secondary antibodies, one labeled with 5 nm of gold and the other with 15 nm of gold. T o avoid misinterpretation caused by possibly different penetration of two sizes of gold particles, in each experiment studies were performed in duplicate, alternating the size of gold particles conjugated with each secondary antibody.

Light Microscopic Immtlnocytocbemistry

Vacuolated muscle fibers of sporadic and hereditary IBM biopsies contained very strongly immunoreactive (IR) inclusions of all PPP sequences within large and small vacuoles (Fig 2) . (For simplicity, the designation IR will be omitted after each substance localized in this study, but that term must be understood.) The PPP-inclusions were located in the subsarcolemma or internally in the fibers. Occasional abnormal muscle fibers, especially ones containing only one or two small vacuoles, had a thin rim of sarcolemmal-like staining (see Fig 2) . In some very abnormal and highly vacuolated muscle fibers, PPP accumulations appeared to extend slightly outside the boundary of the muscle fiber (see Fig 2) . In an occasional vacuolated muscle fiber, in addition to strong immunoreactivity within vacuoles and/or their close proximity, there was also a slight, somewhat diffuse immunoreactivity in what appeared to be vacuole-free cytoplasm with antisera C8 and R37 (both against the C-terminal of PPP) and T97 (against the N-terminal of PPP). In the hereditary IBM biopsy, some muscle fibers contained a strong, wide subsarco-lemma1 accumulation of PPP sequences that extended to the interior of the fiber (Fig 3) . Use of Boehringer antisera against PPP sequences provided the same abnormal pattern of PPP sequence accumulation but with considerably weaker reactivity. Light microscopic double immunostainings showed that (1) Cand N-terminals of PPP closely colocalized with AP (Fig 4A-D) , and (2) that PPP was colocalized with Ub in all PPP-positive fibers (Fig 4E, F) .

When the primary antibody was omitted, absorbed, or replaced by a nonimmune serum, the immunoreaction did not take place.

None Of The Biopsies Of 9 Patients With Polymyositis

and none of the other control patients' biopsies had the PPP inclusions that were characteristic of IBM vacuolated muscle fibers. Small regenerating muscle fibers in IBM, polymyositis, and Duchenne muscular dystrophy had slight, evenly distributed immunoreactivity of Cand N-terminal PPP (with C8, T97, and R37 antisera).

As previously described in normal human muscle 145, 461, in all biopsy categories there was strong immunoreactivity of all PPP sequences and Ub at the postsynaptic domain of the neuromuscular junctions. Intramuscular nerve twigs, in all biopsies that contained them, were immunoreactive with all PPP sequences except AP 8-17 (G-OP-1) and Af3 1-40 (R1280). Table 2 ) Most of the structures identified by all the PPP antibodies consisted of irregular clusters of nearly amorphous dense material, which were either irregular or rounded ( Figs 5 , 6 ). In those structures there was colocalization of all three PPP sequences (see Fig 5) . The three PPP sequences were also colorized as loose, floccular material (see Fig 5 ) , which was often in close proximity to, or adjacent to, the 15to 21-nm CTFs; however, the CTFs themselves never contained immunoreactivity of any of the PPP epitopes (see Fig 6) . In proximity to the loose, floccular material, there were also loosely packed amyloid-like fibrils 6 to 10 nm in diameter; these had AP but not Cor N-terminal PPP immunoreactivity (see Fig 5) .

Ap With 15-Nm Gold Particles. While There Is Close Colocalization Of N -And C-Ppp With Ap On The Denser Amorphous And Looser Floccular Structures Zn A And B, Only A P Is Localized To The 6To 10-Nm Loose Fibrilkzr Structures (B, C). (A:

X 107?000 before 3% reduction; B: X 134,000 before 3% reduction; C:

x 128,000 before 3% reduction.) (0) C-PPP is localized to amorphous structures (5-nm gold particles) ( x 108,000 before

3% Reduction). (E-Gi A@ Is Localized To Amorphous Structures

and fibrillur material extending from the amorphous material (E) and organized in loose rhsters (F, Gj ( I 0-nm gold particles) (E: X 55,000 bq5ore 3% reduction; F , G: X 65,000 before 3% reduction). (H) Intermingled with collagen fibriis lying slightly outside an abnormal muscle fiber are anti-C-PPP antibody immunodecorated structures ( I 0-nm gold particles) ( x 67,000 before 3% reduction). Even though most of the PPP loci were inside the muscle fibers, occasionally immediately outside very abnormal muscle fibers there were also patches of PPP material lying in the extracellular space and intermingled with collagen fibers (see Fig 5) .

Table 2. Subcellular Localization Of Pamyloid Precursor Protein Epitopej-And Ubiquitin In Inclusion Body Myositis Muscle

Discussion

This study demonstrated that N-, C-, and AP epitopes of PPP were abnormally accumulated in the vacuolated muscle fibers of sporadic and hereditary IBM. Most of those accumulations were Congo red and crystal violet positive, indicating amyloid in p-pleated sheets. However, some vacuolated fibers in sporadic IBM and all the vacuolated fibers in our 1 patient with hereditary IBM had three-epitope PPP positive accumulations that were Congo red negative, suggesting that their AP sequences were not in an amyloid P-pleated sheet configuration, but perhaps still in relatively intact PPP molecules. Accordingly, such accumulations may represent early changes of this IBM feature. The Congo red negativity of these A@-containing loci is like that of diffuse plaques in A D brain, which are considered to be an early abnormality {40, 501. The presence of PPP accumulations deeply internal within some muscle fibers suggests their origin was intracellular.

Even though immunoreactivities of PPP epicopes colocalized with each other and with Ub at the light microscopic level, their ultrastructural localization differed. While all three PPP epitopes were colocalized to several types of abnormal subcellular structures, only AP was locahzed to the 6to 10-nm-diameter amyloid-like fibrils. None of the PPP epitopes was localized on the 15-to 21-nm-diameter CTFs; these contained only strong Ub immunoreactivity.

In some very abnormal muscle fibers, the accumulations of PPP appeared to extend outside the muscle fiber boundary. This may have been due to a fragility of the fiber's surface membrane, which could have been transiently broken. Sometimes associated with those fibers were a few PPP structures intermingled with collagen fibrils, either inside the muscle fiber or outside but very close to it. Because collagen fibrils are an extracellular component, the occasional PPP structures intermixed with them just outside the fiber may have been generated (1) inside the muscle fiber but emerged through a broken membrane or by exocytosis, or (2) outside the fiber. The first possibility appears more likely because (1) it has been reported that PPP is synthesized in the normal adult muscle fiber {Sl}, (2) PPP is present within normal human muscle fibers at the postsynaptic domain of the neuromuscular junction Prior to our studies of 1BM muscle, abnormal accumulation of PPP epitopes was reported only in (1) brain senile plaques of Alzheimer's disease and Down's syndrome, and (2) cerebral blood vessels of Alzheimer's disease and Dutch-type hereditary cerebral hemorrhage (reviewed in 122, 27, 281). Our results demonstrate that all three epitopes of PPP (N-terminal, C-terminal, and AP) are abnormally accumulated in vacuolated muscle fibers of sporadic and hereditary IBM. Further studies will be required to answer several questions. Does this accumulation result from increased generation of PPP, abnormal processing of it, or both? Is PPP abnormality the initial step in a cascade of pathogenic events leading to sporadic and hereditary IBM, including formation of CTFs, or is it secondary to another pathogenic mechanism? Is the Ub increase a primary pathogenic process or is it secondary to a need for Ub to facilitate degradation of abnormal proteins? Do similar abnormal PPP accumulations occur in any other muscle disease that we have not yet studied?

The fact that AP, but not C-and N-terminal epitopes of PPP, is localized to the amyloid-like fibrils suggests that free AP is generated during PPP processing and may be responsible for the amyloid present in IBM muscle fibers.

Conclusion

Our study indicates that abnormal accumulation of PPP occurs in diseased human muscle and thus is not unique to the brain and cerebral vessels. Normal and pathological human muscle, including use of cultured normal and genetically abnormal (such as hereditary IBM) muscle [SS}, should provide living human tissue more readily accessible than brain for a wide range of molecular studies of PPP synthesis and processing. Such analyses could lead to a better understanding of sporadic and hereditary IBM, Alzheimer's disease, and perhaps other diseases. Whether there is a mutation of the PPP gene in hereditary IBM remains to be determined.

This study was supported by the Ron Stever Research Fund.

For their generous gifts of anribodies, we are grateful to Drs G. G.

Glenner, D. J. Selkoe, B. Frangione, T. Ishii, and S. Haga. We thank Dr Glenner for synrhetic A@ and Dr Selkoe for synthetic C-PPP epitope.

SECTION

, 1993, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ana.410340408 by University Of Illinois At Urbana Champaign, Wiley Online Library on [07/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

, 1993, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ana.410340408 by University Of Illinois At Urbana Champaign, Wiley Online Library on [07/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

, 1993, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ana.410340408 by University Of Illinois At Urbana Champaign, Wiley Online Library on [07/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

, 1993, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ana.410340408 by University Of Illinois At Urbana Champaign, Wiley Online Library on [07/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

, 1993, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ana.410340408 by University Of Illinois At Urbana Champaign, Wiley Online Library on [07/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

1Y336-342 5 6 0 Annals of Neuroiogy Vol 34 No 4 October 1993 15318249, 1993, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ana.410340408 by University Of Illinois At Urbana Champaign, Wiley Online Library on [07/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License